INDEX

Explanations

phrases indicating a specific event or action occurring in a given location

oai_token-act-pair · gpt-3.5-turbo

the word "the" in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

worn

-0.71

Conn

-0.70

ãĤ´ãĥ³

-0.69

buster

-0.68

Cho

-0.65

peer

-0.63

agree

-0.61

ioned

-0.61

hari

-0.60

=""

-0.60

POSITIVE LOGITS

 sake

1.72

 purposes

1.34

 remainder

1.17

 foreseeable

1.15

 upcoming

1.08

 same

1.08

 purpose

1.06

 aforementioned

0.98

 duration

0.97

 entirety

0.97

Activations Density 0.266%

phrases indicating a specific event or action occurring in a given location

the word "the" in various contexts

No Comments

No Known Activations